Exposure dosimetry surveillance system

Abstract

Apparatus and methods relate to dynamic hazard exposure dosimetry (HED). In an illustrative example, a HED computation engine (HEDGE) may automatically determine a relative three- dimensional (3D) position of a person to a hazard source by analyzing an image feed, such as by a computer vision engine. The 3D position of the person may, for example, be determined by evaluating the size and orientation in the image feed of one or more visual indicia worn by the person and / or disposed on the hazard source. A dosage prediction engine may, for example, generate a hazard dosage based on the relative position over time, such as by using predetermined hazard dosage distributions correlated with the 3D position of the hazard source. Various embodiments may advantageously provide a reduced complexity, easily installed and / or maintained real-time monitoring of hazard exposure (e.g., radiation exposure) in controlled environments (e.g., medical operating rooms).

Description

Exposure Dosimetry Surveillance SystemCORRESPON DENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 566,797, filed on March 18, 2024 by Sean Avery Adams, et al., and titled Exposure Dosimetry Surveillance System.

[0002] This application may relate to subject matter and / or have common inventorship with: US Patent Application No. 18 / 847, 171 filed Sep. 13, 2024, EPO Patent Application No. 23715398.6 filed Sep. 26, 2024, and PCT Patent Application Serial No. PCT / US23 / 64339 filed Mar. 14, 2023, all naming Edward Casteel Milner, et al., and titled “Realtime Hazards Visualization;" and U.S. Provisional Patent Application Nos. 63 / 269,300 titled “Realtime Hazards Visualization” and filed on March 14, 2022, 63 / 362,960 titled “Dynamic Registration of Hazard Visualization” and filed on April 13, 2022, and 63 / 490,159 titled “Realtime Hazards Visualization” and filed March 14, 2023.

[0003] The disclosure of the above-referenced applications and all patents issued therefrom are incorporated herein by reference in their entirety and for all purposes.TECHNICAL FIELD

[0004] Various embodiments relate generally to hazard detection systems.BACKGROU N D

[0005] Dosimetry includes scientific measurements and assessments of doses of ionizing radiation in materials or tissues. For example, dosimetry may be used in radiation protection, medical treatments, environmental monitoring, and nuclear safety.

[0006] In some applications, dosimetry may include, for example, calculation, measurement, and evaluation of radiation doses received by objects or individuals. For example, dosimeters may be used to quantify the absorbed dose of radiation accurately. Dosimeters, for example, may include film badges, thermoluminescent dosimeters (TLDs), optically stimulated luminescence dosimeters (OSLDs), and electronic personal dosimeters (EPDs).

[0007] One popular dosimeter is radiation badges. A radiation badge, for example, may be worn by individuals who work in environments where they may be exposed to ionizing radiation. The radiation badge may, in some cases, include a small piece of radiation-sensitive material (e.g., a photographic film, a thermoluminescent crystal). For example, the radiation-sensitive materialmay absorb radiation energy. By periodically analyzing the badge, for example, through a specialized process, radiation exposure levels of a material attached to the radiation badge may be determined.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 depicts an example Exposure Surveillance and Prediction System (ESAPS) employed in an illustrative use-case scenario.

[0009] FIG. 2A and FIG. 2B are block diagrams depicting example systems for operating and training a dosimetry prediction model.

[0010] FIG. 3 A is a block diagram depicting an example system of a real-time dosimetry computation system.

[0011] FIG. 3B depicts a block diagram of an example Radiation Hazard Warning System.

[0012] FIG. 4 depicts an example user interface displaying radiation accumulation on a person in real-time.

[0013] FIG. 5 is a flowchart illustrating an example data generation process for identifying radiation exposure hazards.

[0014] FIG. 5 is a flowchart illustrating an example user interaction process using an example ESAPS.

[0015] FIG. 6 is a flowchart illustrating an example startup subprocess.

[0016] FIG. 7 is a flowchart illustrating an example operation subprocess.

[0017] FIG. 9 is a flowchart illustrating an example hazard warning subprocess.

[0018] FIG. 10 is a flowchart illustrating an example hazard source (e.g., radiation source) activation detection process.

[0019] FIG. 11 depicts an illustrative wearable, which may, for example, be configured as a personal alert device.

[0020] Like reference symbols in the various drawings indicate like elements.DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

[0021] FIG. 1 depicts an example Exposure Surveillance and Prediction System (ESAPS) employed in an illustrative use-case scenario. For example, the ESAPS may be used in an operation room to measure radiation hazards within the operation room. In this example, an ESAPS 100 includes a real-time dosimetry computation engine (RDCE 105). The RDCE 105 receives signals from a camera 110 (e.g., a digital camera, an optical sensor, a 3-dimensional(3D) camera). The camera 110, in this example, is capturing images of a person 115 under a radiation source 120. In some implementations, the camera 110 may be a Depth Camera D455 of Intel® RealSense™ Depth and Tracking Cameras. Intel® and RealSense™ are trademarks of Intel Corporation headquartered in Santa Clara, California.

[0022] As an illustrative scenario without limitation, the person 115 and the radiation source 120 may be located in an operation room. For example, the person 115 may include a medical personnel (e.g., a surgeon, a nurse). For example, the person 115 may include a patient. For example, the radiation source 120 may include an X-ray source (e g., a C-arm). In some implementations, the radiation source 120 may include other fluoroscopic devices.

[0023] In various embodiments, radiation exposure of the person 115 may be tracked (e.g., in real-time, in near real-time). For example, the person 115 may receive radiation from a direct beam of the radiation source 120. The person 115 may, for example, also receive indirect beam (e g., scattered beam) reflect from other objects (e.g., of the operation room).

[0024] In this example, the person 115 may be wearing one or more tags 125. For example, the one or more tags 125 may include a unique identifier (e.g., a quick response (QR) code) identifying the person 115. Various embodiments are the one or more tags 125 are described in PCT Patent Application Serial No. PCT / US23 / 64339, titled “Realtime Hazard Visualization,” on March 14, 2023, by Edward Casteel Milner, et al., at least some of which inventors are also coinventors of this application. The entire contents of PCT / US23 / 64339 and all priority applications thereto are incorporated herein by reference.

[0025] In various implementations, the RDCE 105 may track radiation exposure in, for example, the operation room, in real-time. In this example, the RDCE 105 includes a computer vision engine (CVE 130) and a realtime dosimetry hazard engine (RDHE 135). The CVE 130 receives measurement signals from the camera 110. For example, within a view of the camera 110, the CVE 130 may generate a 3D location of people (e.g., including the person 115), and 3D location of the radiation source 120 in a room. In some implementations, the CVE 130 may identify an orientation of the person 115 based on joints identified on the person 115. For example, the CVE 130 may identify an orientation of the person 115 based on a predetermined placement of the one or more tags 125 on the person 115. For example, the CVE 130 may generate and track radiation exposure of a skeleton of the person 115 as a function of the one or more tags 125. In someexamples, the CVE 130 may identify the orientation based on machine vision (e.g., trained with an artificial intelligent engine).

[0026] As an illustrative example without limitations, the CVE 130 may determine depth of each objects within a field of view based on the one or more tags 125 within the field of view. For example, during a configuration stage, multiple tags may be attached to the radiation source 120 (e.g., a C-arm). For example, the CVE 130 may determine a distance of the radiation source 120 from the camera 110 based on a known size of the tags 125. In some examples, the camera 110 may include a depth measurement function (e g., a ToF sensor) to determine a position of the person 115 from the camera 110.

[0027] In some implementations, the RDHE 135 may determine a hazard generated by the radiation source 120 within a field of view of the camera 110. For example, the RDHE 135 may generate warnings to the person 115 to reduce radiation exposure of the person 115. The RDHE 135, for example, may receive an object identification result from the CVE 130 to determine a real-time dosimetry measurement on the person 115. For example, the RDHE 135 may generate the real-time dosimetry measurement using a dosage prediction engine (DPE 140). As shown, the DPE 140 includes a dosage prediction model 145. For example, the DPE 140 may apply the dosage prediction model 145 to generate the real-time dosimetry measurement. For example, the dosage prediction model 145 may include a simulation model to generate the real-time dosimetry measurement based on objects in the room (e.g., the operation room), the radiation source 120, identified orientation and position of the person 115 in the room. In some implementations, the simulation model may include a predetermined (e.g., precalculated) radiation model of the room based on an operation to be done. For example, the simulation model may be represented as a heatmap. For example, the simulation model may include a look up table (LUT) of radiation at more than one point in the room when the radiation source 120 is activated. In various examples, the RDCE 105 may be configured to track dosimetry of more than one person within the room. In some examples, the LUT may include coordinates including the person 115 ’ s position and the radiation source 120‘s position, and a corresponding radiation exposure. For example, the DPE 140 may transform the LUT based on the orientation of the room and the person to generate the real-time dosimetry measurements.

[0028] The RDCE 105 in this example further includes user parameters 150. In some embodiments, the DPE 140 may adjust / generate the real-time dosimetry measurement byretrieving and applying the user parameters 150 to the dosage prediction model 145. For example, the user parameters 150 may include historical data 155. The user parameters 150 may, for example, include shielding data 160. For example, the historical data 155 may include historical activities of the person 115. For example, the shielding data 160 may include shielding gear worn by the person 115. For example, the DPE 140 may use the shielding data 160 to calculate the dosage on the person 115 based on the shielding gear worn by the person 115. In some implementations, the shielding data 160 and / or the historical data 155 may be updated automatically by the CVE 130 and / or the RDHE 135. In some implementations, some or all of the user parameters 150 may be entered by a user using a user input device 165. For example, the user may enter the number of surgery and usual gear to be worn using the user input device 165 during a user account setup process.In some implementations, the historical data 155 may an estimated rate of the person 115 (e.g., and other users of the RDCE 105) accumulating radiation. After the RDHE 135 receive the real-time dosimetry measurement from the DPE 140, the RDHE 135, for example, may generate an output to a live visual display 170 and / or a live hazard alarm 175. For example, the RDHE 135 may, based on the real-time dosimetry measurement and the historical data 155 associated with the person 115, generate a visualization of the real-time dosimetry measurement overlaid on an image of the room. For example, the person 115 may advantageously avoid the high radiation area.As an illustrative example, the person 115 may be a cardiologist who regularly performs a heart surgery with a C-arm (e.g., the radiation source 120). For example, the person 115 may habitually be in an orientation during the surgery that may increase the person 115’ s tumor burden on, for example, a left side of the body. In some examples, the RDCE 105 may display one or more visual indicia on the live visual display 170 overlaying on the left side of the person 115. For example, the person 115 may then be more likely to, for example, change a standing position (e.g., and / or orientation) during the surgery to reduce the tumor burden.The live hazard alarm 175 may generate an alarm to warn the person 115. For example, the RDHE 135 may generate an alarm at the live hazard alarm 175 when the person 115 is in a direct beam of the radiation source 120. For example, the RDHE 135 may generate the alarm when the person 115 is exposed to indirect (e.g., back-scattered, reflected) beam for extended period of time.

[0029] In an illustrative example, the CVE 130 may first differentiate between each body part of the person 115 ’ s exposure to the radiation source 120. In some implementations, the RDHE 135 may aggregate radiation by de-projecting a position of the person 115. For example, the CVE 130 may determine pixel locations of the person 115 ’ s joints (e.g., using a matrix translation of a captured image to a computer vision model). For example, the CVE 130 may convert the pixel locations into 3D coordinates in a 3D point cloud. For example, the RDHE 135 may generate a radiation exposure based on the pixel locations. For example, the RDHE 135 may determine whether some or all of the person 115 are within a hazardous radiation zone.

[0030] In some embodiments, the CVE 130 may include a computer vision library (e.g., OpenPose). For example joints of a body may be predetermined in the computer vision library. For example, the computer vision library may be designed for real-time multi-person human pose estimation. In some implementations, the CVE 130 may analyze images or videos and identify the positions of key body points (e.g., joints, hands, and face) on multiple people within the field of view in real-time. For example, the DPE 140 may apply the dosage prediction model 145 (e.g., a LUT of radiation emitted by the radiation source 120 based on the room’s configuration) to determine radiation exposure at the key body points.

[0031] In some implementations, the RDCE 105 may generate a predictive radiation result. For example, the DPE 140 may predict, based on the person 115 ’ s previous history (e.g., retrieved from the user parameters 150) an estimated radiation dosage based on historical activities of the person 115. For example, the estimated radiation dosage may be generated based on a type of procedures scheduled. As shown, the RDCE 105 may generate a time-based exposure report 180 at a user computer 185. For example, the time-based exposure report 180 may include a dosimetry of the person 115 in the past. The time-based exposure report 180 may also include a predicted radiation exposure of the person 115 for, for example, today, this week, this year, based on scheduled activities.

[0032] In various implementations, the time-based exposure report 180 may include suggested changes (e.g., stop leaning in a specific table) to reduce radiation exposure. For example, the time-based exposure report 180 may include a prediction of health risks based on current behavior and improvements based on suggested changes. In some examples, the time-based exposure report 180 may include recommended accessories (e.g., radiation shielding gears) to beused (e.g., “Using leaded goggles would make a huge change based on your practice / behavior.”).

[0033] As an illustrative example, the cardiologist in the previous example may be very focused on surgical operation during each surgery. For example, the cardiologist may lose track of how long that radiation exposure may be on a chest of the patient. For example, the live visual display 170 may be configured to project on the chest. For example, the live visual display 170 may display on a place where the cardiologist may see that the chest has been exposed for over a predetermined period of time and / or calculated radiation dose (e.g., when a time is getting close to a radiation burn level).

[0034] In some implementations, the time-based exposure report 180 may be generated based on radiation exposure to the patient. For example, the RDCE 105 may generate a radiation exposure report to the patient based on a period of time (e.g., in the past and / or projecting into the future).

[0035] For example, the time-based exposure report 180 may include information of the radiation what that radiation can mean for their long term health. In some examples, a medical personnel may advantageously use the time-based exposure report 180 to investigate a reason for a real radiation burn of a patient.

[0036] FIG. 2A and FIG. 2B are block diagrams depicting example systems for operating and training a dosimetry prediction model. As shown in FIG. 2A, a DPE 200 (e.g., the DPE 140 as described with reference to FIG. 1) may receive inputs including procedure, real-time position, real-time orientation, and shielding data and generate a real-time exposure heatmap 205. For example, the real-time exposure heatmap 205 may include a real-time representation of radiation exposure of the person 115 during a selected procedure.

[0037] In some implementations, the RDHE 135 may use the real-time exposure heatmap 205 to generate signal to the live visual display 170 and / or the live hazard alarm 175. For example, the RDHE 135 may, based on an exposure time of a body part of the person 115 under high radiation exposure, generate an alarm signal to the live hazard alarm 175 to notify the person 115 to remove from a dangerous location. In some implementations, the RDCE 105 may store the realtime exposure heatmap 205 periodically (e.g., every minute, every second, every 5 minutes) to generate the historical data 155. Based on the historical data 155, the RDCE 105 may generate the time-based exposure report 180, for example. In some implementations, the RDCE 105 maypredict radiation exposure of the person 115 in a surgery based on historical data (e.g., historical data 155) before the surgery has begun.

[0038] The DPE 200 includes a predetermined radiation exposure LUT 210. For example, the predetermined radiation exposure LUT 210 may include radiation models. For example, the radiation models may include radiation exposure quantity at each point of a point cloud generated by the CVE 130. In some embodiments, the predetermined radiation exposure LUT 210 may be selected based on a procedure being carried out in the operation room. For example, the predetermined radiation exposure LUT 210 may (also) be selected based on the radiation source 120 (e.g., a model of the C-arm and intensity used for a particular surgery). In some examples, the predetermined radiation exposure LUT 210 may include entry based on an orientation of the radiation source 120.

[0039] For example, the CVE 130 may transmit an identified orientation and an identified position of the person 115. For example, based on the predetermined radiation exposure LUT 210, the DPE 200 may generate a real-time exposure of the person 115 at different body parts based on the identified orientation and position. In some examples, the DPE 200 may also discount the radiation exposure at the body part based on the shielding data (e.g., retrieve based on the shielding data 160 of the user parameters 150). As an illustrative example, the person 115 may entered using the user input device 165 of a gender of female. The CVE 130 may, for example, identify locations of a pelvis and neck of the person 115. Accordingly, for example, the RDCE 105 may generate the time-based exposure report 180 based on the gender of the person 115.

[0040] In this example, the DPE 200 also includes a transformation module 215. For example, the DPE 200 may transform (e.g., scale, rotate) the predetermined radiation exposure LUT 210 based on an identified dimension of a room (e.g., based on the field of view of the camera 110). In some implementations, the DPE 200 may quickly generate a distribution of radiation exposure in a room based on a predetermined set of room configurations.

[0041] As shown in FIG. 2B, the DPE 200 is trained by a training engine 220. For example, the training engine 220 may be implemented in the RDCE 105 of FIG. 1. In some embodiments, the training engine 220 may receive predicted dosage data 225 from the DPE 200. For example, the training engine 220 may receive radiation badge data 230 corresponding to a same measurement time period of the predicted dosage data 225. For example, the training engine 220 may generateadjustment signals to the DPE 200 based on a difference between the predicted dosage data 225 and the radiation badge data 230. For example, the DPE 200 may update the dosage prediction model 145 based on the adjustment signal.

[0042] As an illustrative example, the RDCE 105 may be trained for a predetermined period of time (e.g., 1 week, 1 month, 3 months). For example, the person 115 may wear the radiation badge data 230 on various part of the body. For example, using the RDCE 105, the ESAPS 100 may also generate the predicted dosage data 225 based on calculations from the RDHE 135. For example, when an error (e.g., a difference) between the radiation badge data 230 and the predicted dosage data 225 is larger than a predetermined threshold, the training engine 220 may generate a signal to adjust parameters of the dosage prediction model 145 to minimize the error.

[0043] In various implementations, the dosage prediction model 145 may include the predetermined radiation exposure LUT 210. For example, the training engine 220 may update entries in the predetermined radiation exposure LUT 210 to minimize the error between the predicted dosage data 225 and the radiation badge data 230.

[0044] FIG. 3A is a block diagram depicting an example system of a real-time dosimetry computation system (RDCE 105). The RDCE 105 includes a processor 305. The processor 305 may, for example, include one or more processing units. The processor 305 is operably coupled to a communication module 310. The communication module 310 may, for example, include wired communication. The communication module 310 may, for example, include wireless communication. In the depicted example, the communication module 310 is operably coupled to the live visual display 170, the user input device 165, the camera 110, and a hazard notification device 315. For example, the hazard notification device 315 may include the live hazard alarm 175. For example, the hazard notification device 315 may include other notification devices. For example, the hazard notification device 315 may include one or more visual indicia (e g., a tower light). For example, the hazard notification device 315 may include a mobile device that generate warning messages / popup when a hazard is detected.

[0045] The processor 305 is operably coupled to a memory module 320. The memory module 320 may, for example, include one or more memory modules (e.g., random-access memory (RAM)). The processor 305 includes a storage module 325. The storage module 325 may, for example, include one or more storage modules (e.g., non-volatile memory). In this example, thestorage module 325 includes the CVE 130, the RDHE 135, the DPE 140, and the training engine 220.

[0046] The processor 305 is further operably coupled to a data store 330. The data store 330 includes the user parameters 150. For example, the user parameters 150 may include the historical data 155 and the shielding data 160. In some implementations, the user parameters 150 may also include user identification information. For example, the user parameters 150 may include log in information of the person 115. For example, the RDCE 105 may store activity (e.g., procedure) history of the person 115 associated with the user identification information.

[0047] The data store 330 also includes the predetermined radiation exposure LUT 210. For example, the DPE 140 may retrieve the predetermined radiation exposure LUT 210 to generate the real-time exposure heatmap 205 in operation. In this example, the data store 330 includes a historical exposure heatmap 335. For example, the RDCE 105 may store the real-time exposure heatmap 205 generated by the DPE 140 to the data store 330. For example, the RDCE 105 may associate the historical exposure heatmap 335 to a corresponding user. In some examples, the DPE 140 may apply the dosage prediction model 145 in the data store 330 to the historical exposure heatmap 335 to generate the time-based exposure report 180 for each user of the ESAPS 100.

[0048] FIG. 3B depicts a block diagram of an example Radiation Hazard Warning System (RHWS 300). As shown, the RHWS 300 includes a 3D camera 340. For example, the 3D camera 340 may include the camera 110. For example, the RHWS 300 may be used in a location for identifying hazardous situations based on measurements from the 3D camera 340.

[0049] In this example, the RHWS 300 includes a 3D position / orientation identification engine (TPOIE 345). For example, the TPOIE 345 may process images captured by the 3D camera 340 (e.g., camera 110) For example, the TPOIE 345 may include a pose identification model (e.g., OpenPose models). For example, the TPOIE 345 may generate a pixel location result including pixel locations of every node for every person in a field of view of the 3D camera 340 (e.g., during a single frame). For example, the pixel locations may be represented in this form: (x pixel location, y pixel location, confidence score).

[0050] In some implementations, the TPOIE 345 may process the pixel locations by generating a set of 3D point locations for every node. In some examples, when a number of people are in the field of view, each person may include a predetermined node (e.g., 25 points) associated with abody’s location in 3D space. In some embodiments, the TPOIE 345 may generate the 3D points in a format of (e.g., in meters): (x distance from point in center of camera’s view, y distance from point in center of camera’s view, depth from camera).

[0051] As shown, the RHWS 300 includes a radiation engine 350 and hazard engine(s) 355. For example, the radiation engine 350 and the hazard engine(s) 355 may be configured to generate potential hazard exposure information of one or more person in the field of view. For example, the radiation engine 350 may include the RDCE 105. In some implementations, the hazard engine(s) 355 may include an engine to identify potential violent person in a premises. For example, the hazard engine(s) 355 may include an engine to identify potential fire hazards.

[0052] In some implementations, the hazard engine(s) 355 may, based on the predetermined nodes associated with the body of each person, track a hazard exposure for each body part. In some examples, the radiation engine 350 and / or the hazard engine(s) 355 may use a mapping between a real-time hazard exposure and a location within the field of view (e.g., the dosage prediction model 145, the predetermined radiation exposure LUT 210).

[0053] In this example, after the hazards are identified by the radiation engine 350 and / or the hazard engine(s) 355, the RHWS 300 may generate a prediction result 360 or a hazard display 365. For example, the prediction result 360 may be generated at the user computer 185. For example, the prediction result 360 may be delivered to the user upon request (e.g., after login with user’s credentials). In some examples, the hazard display 365 may be displayed (e.g., in real-time) on the live visual display 170.

[0054] FIG. 4 depicts an example user interface displaying radiation accumulation on a person in real-time. As shown, an actual radiation exposure user interface (AREUI 400) may generate a visual display of radiation exposure of a room 405. For example, the AREUI 400 may be displayed on the user input device 165. For example, the user input device 165 may include a mobile device. For example, the AREUI 400 may be generated by the live visual display 170. For example, the AREUI 400 may include an overlaying image of various radiation exposure rate (e.g., determined by the DPE 140 using the dosage prediction model 145) on a dynamic image captured by the camera 110.

[0055] In this example, For example, a person 410 is wearing a tag 415. In some implementations, tags 420 may also be applied to equipment in the room 405. For example, the CVE 130 may identify the tag 415 and the tags 420 to determine positions (e.g., along alatitudinal axis, a longitudinal axis, and a depth) and orientation of the user and other fixtures within the room 405. Using the AREUI 400, a user may advantageously recognize radiation accumulation locations within the room 405.

[0056] As shown, the AREUI 400 includes user input 425. For example, a user may use the user input 425 to select whether a C-arm is on or off. For example, the RDCE 105 may generate the AREUI 400 based on the user inputs. For example, when the C-arm is off, the AREUI 400 may transmit a signal to the RDCE 105 to stop calculating radiation accumulation within the room 405.

[0057] In this example, a visual hazard overlay 430, depicted on a live visual display 170, represents the distribution of radiation exposure within a monitored environment, such as a room 405. The visual hazard overlay 430 may advantageously allow a person 115, depicted as wearing a tag 415 (e.g., one of the tags 125), to visualize radiation hazard areas.

[0058] The hazard overlay 430 may be generated by retrieving a predetermined radiation distribution, as shown in FIG. 4. This process may, for example, include accessing data corresponding to known radiation patterns (e.g., specific to the radiation source and / or environment). Once retrieved, the predetermined radiation distribution may, for example, be translated, scaled, and / or rotated to align with the radiation source.

[0059] Such adjustments may, for example, be based on the orientation of the radiation source 120. The orientation of the radiation source 120 may be determined using real-time monitoring of visual indicia, such as the tags 420 (e.g., which may be tags 125). By detecting the size and placement of tags 420, the system may, for example, use the camera 110 to determine the exact orientation and position of the radiation source 120 within the monitored environment.

[0060] This dynamically adjusted radiation distribution may, for example, be used to generate the visual hazard overlay 430 by transformation (e.g., rotate, translate, scale). The overlay may, for example, advantageously provide real-time visual feedback regarding radiation intensity patterns, thereby enabling individuals to visualize potential exposure risks relative to their positions within the room 405.

[0061] In this example, the hazard overlay provides visual cues of levels of hazard intensity. Multiple indicators may, for example, represent varying levels of radiation intensity across the room 405. For example, first hazard zone 435 may correspond to a first level of intensity (e.g.,lowest displayed level). Second hazard zone 440 may, for example, correspond to a second level of intensity. Third hazard zone 445 may, for example, correspond to a third level of intensity.

[0062] In this example, the hazard overly 430 includes a reflected hazard zone 450. The reflected hazard zone 450 may, for example, advantageously illustrate radiation scatter patterns.

[0063] The hazard overlay 430 may, for example, advantageously assist users in making informed decisions to modify behavior and positioning to reduce exposure.

[0064] FIG. 5 is a flowchart illustrating an example data generation process for identifying radiation exposure hazards. For example, the RDCE 105 may use a data generation process 500 to generate radiation accumulation results for each person identified in an operation room. For example, the CVE 130 may include a computer vision library to operate the data generation process 500. In this example, the data generation process 500 begins when visual data is received from a 3D camera in step 505. For example, the camera 110 may capture (3D) images from an operation room of the person 115 and the radiation source 120.As shown, visual data may be passed to a pose estimation process (PEP 510). The PEP 510 may, for example, utilize an OpenPose model. For example, the PEP 510 may generate pixel locations of each person’s joints. For example, the CVE 130 may identify pixel locations of the person 115 based on a computer vision library. After getting the pixel locations, the data generation process 500 includes a deprojection process 515. For example, the deprojection process 515 may generate 3D locations of each person’s joint from the pixel location of each person’s joint. In some implementations, the data generation process 500 may generate all person’s 3D location data 520. For example, the CVE 130 may generate a data fde of all person’s 3D location data within a field of view of the camera 110.

[0065] In a parallel process, in some implementations, the data generation process 500 includes a visual indicia recognition process (VIRP 525). The VIRP 525 may identify locations and patterns of visual indicia within the visual data. As an example, ArUco tags may represent a type of visual indicia, which may be wearable, and may include a unique pattern to identify a person or fixture associated with each of the one or more visual indicia (e.g., tags 125). Based on such visual indicia, the VIRP 525 may determine person identities 530 in the visual data. The VIRP 525 may also determine radiation source location process 535 using the visual data.

[0066] In this example, the data generation process 500 includes a check relative locations process 540. For example, the check relative locations process 540 may determine a real-timeradiation exposure for each person identified in the visual data based on a geometrical relationship between the radiation source determined in radiation source location process 535 and the persons 3D location data 520. For example, the RDHE 135 may use the DPE 140 to generate, for each person 115 in a field of view of the camera 110, a real-time exposure data.

[0067] After the real-time radiation exposure is determined, the data generation process 500 includes an accumulated radiation on screen via a correct person process 545, and the data generation process 500 ends. The correct person process 545 may, for example, reassign accumulated radiation levels by correlating a visual indicia identified in the VIRP 525 with the person detected in the 3D location data 520, relative to the radiation source location process 535, such as determined during the check relative locations process 540. For example, the RDHE 135 may use the check relative locations process 540 to update the live visual display 170.

[0068] FIG. 6 is a flowchart illustrating an example user interaction process 600 using an example ESAPS. For example, the RDCE 105 may receive input and response to the user input using the process 600. In this example, the user interaction process 600 begins with a startup subprocess 700 (described with reference to FIG. 6). After the startup subprocess, the process continues with an operation subprocess 800 (described with reference to FIG. 7). At an end of the operation subprocess 800, it is determined whether a hazard warning is to be generated in step 605. If no warning is to be generated, it is determined whether the user interaction process 600 is to be ended. If the process is not to be ended, the startup subprocess 700 is followed. If the user interaction process 600 is to be ended, the user interaction process 600 ends.

[0069] In the step 605, if a hazard warning is to be generated, a hazard warning subprocess 900 (described with reference to FIG. 9) is performed. After the hazard warning subprocess 900, the operation subprocess 800 is performed.

[0070] FIG. 7 is a flowchart illustrating an example startup subprocess 700 of the user interaction process described with reference to FIG. 5. For example, the startup subprocess 700 may be performed by the RDCE 105 based on input received from the user input device 165. In this example, the startup subprocess 700 begins when authentication (e.g., a login information is entered) is performed in a step 705. For example, a user (e.g., the person 115) may enter user credentials using the user input device 165. In a decision point 710, it is determined whether the login is correct. If the login is not correct, the step 705 is repeated. If the login is correct, in step 715, it is determined whether any users and visual indicia (e.g., oriented visual indicia, such asArUco tag) associations are to be added. If one or more users and visual indicia associations are to be added, in step 720, a visual indicia (e.g., ArUco) tag is scanned. For example, the visual indicia may be scanned by a camera (e.g., the camera 110, a smart phone camera, a digital camera). Next, a personal identification (ID) is assigned to the scanned visual indicia in step 725, and the step 715 is repeated.

[0071] If no more users and ArUco tag associations are to be added, in a decision point 730, it is determined whether an exit button is pressed. For example, the exit button may be embedded in a user interface. For example, the exit button may be embedded in the AREUI 400. For example, the exit button may be embedded in an equipment (e.g., the radiation source 120, a C-arm). If the exit button is pressed, the startup subprocess 700 ends.

[0072] If the exit button is not pressed, the startup subprocess 700 determine whether C-arm setting is correct in step 735. For example, the startup subprocess 700 may generate a preview message to ask a user to confirm the C-arm setting. If the C-arm setting is not correct, the C-arm is scanned in step 740 and the step 735 is repeated.

[0073] If the C-arm setting is correct, in step 745, it is determined whether a peak hazard setting (e g., a kilovoltage peak (KVP) setting) is correct. For example, the KVP may include a peak voltage applied to the radiation source 120. For example, the KVP setting may determine an energy of the C-arm related to radiation dose. If the KVP setting is not correct, the KVP setting is adjusted in step 750 and the step 745 is repeated.

[0074] If the KVP setting is correct, in step 755, it is determined whether collimator settings are correct. For example, a collimator may control a shape and size of an emission from the radiation source 120. If the collimator settings are not correct, the collimator settings are adjusted in step 760 and the step 755 is repeated.

[0075] If the collimator settings are correct, in step 765, it is determined whether alarm settings are correct. For example, an alarm setting may include one or more threshold, when exceeded, generate a warning alarm to the user. If the alarm settings are not correct, the alarm settings are adjusted in step 770 and the step 765 is repeated. If the alarm settings are correct, the startup subprocess 700 ends and proceed to operation subprocess 800.

[0076] FIG. 8 is a flowchart illustrating an example operation subprocess 800. The operation subprocess 800 begins in step 805 to determine, for example, whether a radiation source (e.g., C- arm) is on. For example, the C-arm is on when the operation subprocess 800 receives a signalfrom the user (e.g., using the AREUI 400) indicating that the C-arm is on. For example, the C- arm is determined to be on when the operation subprocess 800 receives a signal from the radiation source 120. If the C-arm is not on, the startup subprocess 700 is performed. If the C- arm is on, in step 810, users are tracked and determined. For example, the CVE 130 may track and determine user with a field of view of the camera 110.

[0077] In various embodiments, the detection that the radiation source 120 is on may be automatically determined. Such embodiments may, for example, provide advantageous versatility in various operational environments.

[0078] For example, the RDCE 105 may monitor the movement of an image captured by the camera 110. For example, movement of the image may be detected (e.g., by comparing locations and / or objects within the image over time) wherein the movement may indicate operations occurring ahead of and / or during radiation emission, such as a person 115 breathing. The detected movement may, for example, be used by the RDCE 105 to generate a signal to activate radiation detection (e.g., people are in the room, a patient is present) and / or infer pending activation of the radiation source 120.

[0079] In some examples, the RDCE 105 may detect the activation of the radiation source 120 by identifying changes in visual indicia (e.g., indicator lights). The CVE 130 may, for example, be operated to detect specific illuminations on the radiation source 120 and / or on an associated control panel corresponding to radiation source 120 activation status.

[0080] In some implementations, for example, the RDCE 105 may be configured to utilize the detection of user interaction with radiation activation controls. For instance, the RDCE 105 may capture images and / or data signals indicating an operator (for example, a person detected as a physician, a radiation technician) pressing an emit button and / or other controls associated with emission by the radiation source 120 in a radiological environment.

[0081] In an illustrative example, the RDCE 105 may be operably coupled to auditory (e.g., microphone(s)) and / or vibration (e.g., vibration sensor, detection of vibration in image feed) sensing modules to detect operational sounds and / or vibrations emanating from a radiation source 120 and / or associated device(s). For example, the RDCE 105 may be configured to recognize activation of the radiation source 120 when a noise associated with the radiation emission process (e.g., the hum of a radiation generator), indicating that the radiation source 120 is operational and emitting.

[0082] Next, in decision point 815, it is determined whether any user is in a backscatter cone. For example, the operation subprocess 800 may receive data from the check relative locations process 540 to check whether any user is in the backscatter cone. If no one is in the backscatter cone, the step 805 is repeated. If anyone is in the backscatter cone, the cone (e.g., radiation intensity distribution) is determined in step 820. For example, the DPE 140 may associate the cone with a predetermined radiation exposure. In step 825, a time is counted. For example, a duration of the user staying within the cone is stored. In step 830, the operation subprocess 800 applies weighting factors. For example, the shielding data 160 may be applied. Next, radiation accumulation is incremented to the user in step 835, and the hazard warning subprocess 900 is performed. For example, the radiation accumulation generated as a function of the cone, the time, and the weighting factors are updated to the historical data 155 associated with the user.

[0083] After the step 810, the operation subprocess 800 also includes determining whether a user is in a direct beam of a radiation source in decision point 840. For example, a surgeon may be under a direct beam of a C-arm while operating on a patient. If the user is not in a direct beam of a radiation source, the step 805 is repeated. If the user is in a direct beam of a radiation source, the step 820 is performed and an alarm is played in step 845. For example, the live hazard alarm 175 may be activated. Next, in decision point 850, it is determined whether the user is still in the direct beam. For example, the CVE 130 may determine whether the user is still in the direct beam using a computer vision library. If the user is still in the direct beam, the step 845 is repeated. If the user is not in the direct beam, the step 805 is repeated.FIG. 9 is a flowchart illustrating an example hazard warning subprocess 900. In this example, the hazard warning subprocess 900 begins after the operation subprocess 800 (e.g., after the step 835. In a decision point 905, it is determined whether a time of a user exposed to a radiation (e.g., including under backscatter and direct beam) is more than a first predetermined threshold (e.g., 10 seconds, 30 seconds, 1 minute). If the time of the user exposed to the radiation is less than the first predetermined threshold, it is determined whether a second alarm is played to the user in decision point 910. If the second alarm is played, the hazard warning subprocess 900 returns to the operation subprocess 800. If the second alarm is not played, the second alarm is played for 3 seconds in step 915. For example, the live hazard alarm 175 may play the second alarm for 3 seconds. In step 920, an alarm flag is set to true, and the hazard warning subprocess 900 returns to the operation subprocess 800.

[0084] If the time of the user exposed to the radiation is more than the first predetermined threshold, in a decision point 925, it is determined whether the time of the user exposed to the radiation is between the first predetermined threshold and a second predetermined threshold. If the time of the user exposed to the radiation is between the first predetermined threshold and the second predetermined threshold, it is determined whether a third alarm is played to the user in decision point 930. If the third alarm is not played, the third alarm is played for 3 seconds in step 935. For example, the live hazard alarm 175 may play the third alarm for 3 seconds. In step 940, a second alarm flag is set to true, and the hazard warning subprocess 900 returns to the operation subprocess 800.

[0085] If the time of the user exposed to the radiation is more than the second predetermined threshold, in a decision point 945, it is determined whether the time of the user exposed to the radiation is between the second predetermined threshold and a third predetermined threshold. If the time of the user exposed to the radiation is between the second predetermined threshold and the third predetermined threshold, it is determined whether a fourth alarm is played to the user in decision point 950. If the fourth alarm is not played, the fourth alarm is played for 3 seconds in step 955. For example, the live hazard alarm 175 may play the third alarm for 3 seconds. In step 960, a third alarm flag is set to true, and the hazard warning subprocess 900 returns to the operation subprocess 800.

[0086] If the time of the user exposed to the radiation is more than the third predetermined threshold, a fifth alarm is played for 3 seconds in step 965, and the hazard warning subprocess 900 returns to the operation subprocess 800.FIG. 10 is a flowchart illustrating an example hazard source (e.g., radiation source) activation detection process. The process initiates with step 1005, in which an RDCE (e.g., the RDCE 105) is initialized. For example, the RDCE 105 may be initialized to detect signals within the operating environment. The RDCE monitors, in step 1010, the image feed from the sensor(s) (e.g., camera 110, acoustic sensor, data communication ports(s)).

[0087] In step 1015, a training phase is initiated. The training phase may, for example, include analyzing the image feed to detect changes correlating to a change in activation state of the radiation source. An operator indication of radiation source activation change is received in a step 1020. The operator indication may, for example, advantageously enable the RDCE toconfirm the activity state of the radiation source. For example, the RDCE may correlate changes in the image feed to the activity state.

[0088] Upon activation of the radiation source, signals from the sensor(s) are recorded (e.g., ongoing from step 1010) in step 1025. Once it is determined in step 1030 that the operator has indicated an activation change (e g., radiation source changed from on to off), the method (e.g., by RDCE 105) proceeds to detect, in step 1035, signal changes in the recorded signal data (e.g., image feed recorded). The detected changes are correlated, in a step 1040, with the changes detected in step 1035.

[0089] In step 1045, an activation detection model is trained based on the training data generated and correlated in the previous steps. The training may, for example, include separating the training data set into a training set and a testing set. The model may be trained, for example, using the training set. The resulting model may, for example, be tested using the testing set. This process may be repeated (e.g., with additional training and / or testing data) until a training parameter (e.g., accuracy threshold) is reached. In this example, the model (e.g., trained, during training) is applied to the sensor data (e.g., enriched as in steps 1035-1040) to categorize, in step 1050, the sensor data into distinctive patterns corresponding to the radiation activation state(s)

[0090] Once it is determined in step 1055 that training is complete (e.g., after a predetermined training sequence is completed, upon receiving a termination of training by the operator, and / or once a predetermined accuracy threshold is reached), then the trained model (e.g., including signal profiles, as depicted) may be used, as in step 1060, to detect radiation emissions based on activation of the radiation source as a function of signals received.

[0091] The activation detection model(s) may, for example, be trained to detect one or more sensory changes correlated to activation detection. The detection of the activation status of a radiation source, such as radiation source 120, may be performed, for example, via utilization of image feed data from camera 110. The image feed may, by way of example and not limitation, be analyzed for changes in brightness levels. A change in brightness level may, for example, correlate with a change in activation status of the radiation source 120. A significant increase in brightness (e.g., of a region corresponding to a light operated during emission) may, for example, indicate that the radiation source is active.

[0092] Some examples may, for example, perform identification of illuminated indicator lights on the radiation source 120 itself. Such lights may, for example, serve as operational signals todenote whether the radiation source is emitting. A computer vision engine, such as CVE 130, may be configured, for example, to detect the state of these lights using image analysis.

[0093] In some embodiments, the activation detection model may be configured to use changes in the shadow or outline of the radiation source 120 captured in the image feed to detect activation status. Variations in the shadow or outline characteristics may, for example, indicate movement and / or an operational status change in the radiation source.

[0094] In some examples, the model may be trained to detect user interaction with the control panel or elements associated with the radiation source 120. Image feed data may, for example, be processed to identify gestures or movements by the user (e.g., that typically precede and / or accompany the activation of the radiation source).

[0095] Visual indicia such as tags 125 (e.g., which may be affixed to the radiation source 120) may be configured to convey operational status. For example, the model may be trained to detect tags and / or other visual cues visible in the image feed to denote the activation status of the radiation source.

[0096] Identification of movement within the image feed may, for example, correspond with activation of the radiation source, such as changes in position of operating personnel or apparatus alignment with respect to the radiation source. The model may, for example, be trained to detect such changes as a signal for activation detection.

[0097] FIG. 11 depicts an illustrative wearable device 1110, configured as a personal alert device. The wearable device 1110 is shown in operational proximity to a user, such as the depicted example worn on a limb 1105 of a person 115. The wearable device 1110 may, for example, include an alert system operable to provide notifications related to hazard exposure levels. The notifications may, by way of example and not limitation, be communicated to a user through various means. Communication means may, for example, include visual, auditory, and / or tactile outputs.

[0098] The wearable device 1110 may include, for example, an integrated alert mechanism. For example, the depicted embodiment includes an array of selectively operable visual indicium 1120, including LEDs (as shown), in some embodiments, the wearable device 1110 may, for example, include a screen display. For example, interface 1115 is shown as a cover, but may include a screen. In the depicted example, the interface 1115 may, for example, include one or more regions that are at least partially translucent. For example, the alert mechanism mayilluminate in differing color patterns corresponding to varied levels of received hazard dosage. For instance, such patterns may include colors such as green, yellow, and / or red to visually indicate exposure categories ranging from safe to critical thresholds. This visual feedback may, for example, advantageously provide users a direct, substantially real-time intuitive indication of radiation exposure status.

[0099] The wearable device 1110 may, for example, incorporate an auditory feature. For example, a speaker module may be included and configured. The speaker module may, for example, be operated to emit sound alerts when specific radiation levels are attained or exceeded. Such sound alerts may, for example, serve as an effective warning for users, including instances when they are not visually engaged with the wearable device.

[0100] The wearable device 1110 may, for example, include haptic feedback components, such as vibration motors. The haptic feedback components may, for example, provide tactile notifications directly to the user. This feature may, for example, be beneficial for ensuring awareness, such as in environments that are noisy and / or when visual alerts may be ineffective.

[0101] In some examples, the wearable device 1110 may interface with a central computer system (e.g., the RDCE 105) as shown in Fig. 1. The wearable device 1110 depicted includes a communication module 1125 (e.g., a radio module). The communication module 1125 may enable use of wireless communication protocols such as, by way of example and not limitation, Bluetooth Low Energy (BLE), to communicate data to and / or from the central computer system. By integrating with a central system, the wearable device 1110 may, for example, receive and / or send information dynamically, allowing for real-time updates and adjustments to notifications based on evolving conditions within the environment. Such embodiments may, for example, advantageously enable the wearable device 1110 to be relatively inexpensive, such as by receiving indications of hazard level without requiring on-board detection and / or computation.

[0102] The wearable device 1110 may, such as in the depicted example, include wired communication modules. A wired interface 1130 (e.g., USB such as USB-C) port may, for example, enable data interchange and / or power coupling (e.g., charging). In some embodiments, a wireless charging module may be included.

[0103] Each wearable device 1110 may, for example, be registered with (e.g., with the RDCE 105). Such embodiments may, for example, advantageously enable precise tracking and monitoring of individual exposure levels across various users and environments. In someexamples, a registration process may involve, for example, associating the wearable device 1110 with specific visual indicia (e.g., one or more tags 125) worn by the user of the wearable device 1110. For example, a person 115 may show the wearable device 1110 and their one or more tags 125 to the camera 110 (e.g., prior to beginning operation of a radiation source 120). In some embodiments, a visual indicium (e.g., one or more tags 125) may be embodied (e.g., applied) on the wearable device 1110. In some examples, the wearable device 1110 may provide the visual indicium for the person 115 (e.g., without needing one or more additional tags 125).

[0104] Although various embodiments have been described with reference to the figures, other embodiments are possible. In some implementations, the RDCE 105 and the camera 110 may be integrated into a regular C-arm. For example, a radiation (e.g., the X-ray) data may be directly obtained from the radiation source 120 via a data cable. For example, the live visual display 170 may be coupled to the C-arm to receive the AREUI 400.

[0105] Although an example system has been described with reference to FIGS. 1-3B, other implementations may be deployed in other industrial, scientific, medical, commercial, and / or residential applications. In some implementations, the RHWS 300 may be used to monitor movement efficiency of a process (e.g., an operation, a factory floor manufacturing process). For example, the RHWS 300 may generate recommendation based on observation on people’s movement within a field of view of the 3D camera 340. For example, the RHWS 300 may be used in an airstrip. For example, the engine(s) 355 may be configured to identify personnel positions near airplanes. For example, the engine(s) 355 may generate a display to warn the personnel to avoid high risk area (e.g., avoid being sucked into engines). So having like that 3D display, especially they could

[0106] In some implementations, the 3D camera 340 may include a drone. For example, the hazard engine(s) 355 may include an industrial hazards engine. For example, the drone may be used to monitor a (e.g., outdoor, a battlefield) working environment.

[0107] In some implementations, the RHWS 300 may be implemented into a car. For example, the engine(s) 355 may be configured to predict a hazard and display the hazard to a user. For example, the hazard engine(s) 355 may include a carjacking engine. For example, the caijacking engine may receive an input from a camera of a surrounding of a car and identify a legitimate user of the car. For example, the caijacking engine may output a hazard display indicating that the user is going (e.g. or having a high probably be) carjacked because someone is approachingthe vehicle. For example, the hazard display may generate a warning signal when somebody is coming up behind the user. For example, the carjacking engine may generate the warning signal (e.g., only) when the user is unlocking the car.

[0108] In some implementations, the hazard engine(s) 355 may be configured to define danger zones around a house. For example, the hazard engine(s) 355 may generate hazard signals to a user when activity is detected in the danger zones within a predetermined time (e.g., between 9pm and 6am).

[0109] In some implementations, the RHWS 300 may be configured to monitor whether medical personnel actually wash their hands before a surgery. For example, the hazard engine(s) 355 may be configured to identify that water and soap are both applied to the medical personnel’s hand for a predetermined time before clearing the medical personnel for surgery.

[0110] In various examples, a RDCE 105 may be coupled to a single camera. The use of a single camera 110, as depicted in FIG. 1, may, for example, advantageously reduce the complexity and cost associated with multi-camera setups. This approach may, for example, simplify installation and maintenance processes due to the reduced hardware requirements. Employing a single camera for monitoring tasks may, for example, advantageously provide streamlined data processing workflows (e.g., the system may only need to handle inputs from a single imaging device). By leveraging a single camera, the system may also achieve enhanced spatial efficiency, enabling implementation in environments where space constraints do not permit larger setups. The use of visual indicators (e g., wearable symbols of known geometry such as determinate orientation and size) and / or integration of advanced computer vision algorithms may, for example, allow a single camera to effectively track real-time exposure levels and user positions relative to hazards, as processed by the CVE 130.

[0111] In various examples, the real-time response of the RDCE 105 may include calculations and updates occurring on the order of minutes, seconds, and / or sub-seconds (e.g., micro-second, millisecond), as shown in FIG. 4. For instance, the DPE 140 may calculate a radiation dosage every minute to provide updated data to the live visual display 170. This may advantageously allow for timely adjustments by the user based on accumulated exposure.

[0112] The CVE 130 may, for example, process and analyze the image feed to determine the user’s spatial position relative to the radiation source 120, updating these calculations every second. Continuous update may, for example, advantageously enable the system to quicklyreflect any movements by the user, which may be crucial during dynamic operations, such as medical procedures.

[0113] In sub-second intervals, the RDHE 135 may, for example, activate the live hazard alarm 175 effectively immediately if threshold exposure levels are detected. Sub-second response may, for example, advantageously provide practically immediate notification to the user. Sub-second responses may be crucial in preventing prolonged exposure in rapidly changing and / or high- hazard (e.g., high radiation intensity) environments.

[0114] In some embodiments, a kit may including various components, such as may be useful for implementing a radiation dosimetry system. For example, a kit may include one or more camera 110 configured for capturing an image feed. The kit may, for instance, include mounting hardware suitable for affixing the camera within an operational environment.A kit may include one or more visual indicium tags 125. One or more of the tags 125 may, for example, be configured to be worn by a person 115. The tags may include, by way of example and not limitation, quick response (QR) codes and / or ArUco markers designed for ease of detection by the camera 110. A kit may, for example, include wearable items, such as clothing and / or personal protective equipment (e.g., radiation blocking such as lead protective garments). Tags 125 (e.g., tags 415) may, for example, be applied to one or more of the wearable items (e.g., lead apron, coat).

[0115] In some embodiments, a kit may include one or more wearable device 1110. The wearable device may, for example, be equipped with pre-installed tags 125 (e.g., tags 420). These wearables may facilitate the detection and identification of individuals and / or other objects within a monitored environment.

[0116] A kit may include, for example, predetermined visual indicium (e.g., of the tags 125) designed (e.g., as a sticker) to be applied to a hazard source (e.g., radiation source 120). These predetermined visual indicium may, for example, provide data used for assessing the spatial position of the hazard source within the camera's field of view.A kit may, for example, include a computer. The computer may, for example, include pre-installed software configured for real-time dosimetry computation (e.g., one or more components of the RDCE 105). The software, for example, may be pre-configured to execute a DPE 140, engage in image processing using a CVE 130, and / or determine dosage outcomes as a function of the captured image feed.

[0117] Some kits may, for example, be configured for installation in diverse environments, including medical operating rooms. Such kits may, for example, advantageously enhance safety measures by providing easily installed hardware and / or software configured to provide immediate feedback and real-time dosimetry data to users.

[0118] Some embodiments may be configured for other than human bodies to monitor and / or manage radiation exposure. For example, livestock within an agricultural setting may be observed using the RDCE 105. For example, the RDCE 105 may be configured to enhance health and / or safety in environments with hazard exposure sources, such as radiation-emitting veterinary diagnostic imaging equipment. In this context, an animal may, for example, be fitted with a visual indicium of predetermined geometry, such as described with respect to the tags 125. Such embodiments may advantageously enable accurate tracking and assessment of radiation exposure.

[0119] As an illustrative example, a veterinary diagnostic suite where a livestock animal, for example a cow, may be under examination using a radiographic device, may be equipped with an ESAPS 100. By applying a dynamic dosimetry computation engine, for example, equipped with a camera to capture real-time images, the relative three-dimensional position of the livestock to the radiation source may be determined. The camera data may, for example, be analyzed to identify the cows' spatial orientation and exposure level. A dosage prediction engine (e.g., DPE 140), may generate a radiation dosage model, allowing caretakers to adjust animal positioning and / or employ shielding strategies to optimize exposure settings.

[0120] In an illustrative example, a ESAPS 100 may be configured for wildlife monitoring, such as in conservation areas subject to radiation from research or industrial activities. For instance, a large mammal such as an elephant may be fitted with tracking tags. These visual indicia may, for example, enable continuous position monitoring relative to a radiation hazard source. Managers may, for example, utilize real-time dosimetry data to create safety protocols and / or reposition animals to areas of lower risk. Such adaptation may, for example, enhance ecological safety of flora and / or fauna in radiation-prone environments by employing predictive computing models akin such as by the RDCE 105.

[0121] In some embodiments, an ESAPS 100 may be deployed for machines (e.g., robot, such as used in radiation environments. For example, a RDCE 105 may be configured to monitor humans (e.g., military personnel), animals, and / or autonomous drones monitoring radiationlevels in post-nuclear incident zones. Fitting machines with recognizable visual indicia, automated systems may, for example, be configured to manage the robots' movements, managing their functionality while maintaining operational safety through real-time exposure evaluation. Such embodiments may, for example, advantageously enable response in dynamically changing environments without human intervention, which may, for example, be crucial in high-dose areas.

[0122] In various embodiments, some bypass circuits implementations may be controlled in response to signals from analog or digital components, which may be discrete, integrated, or a combination of each. Some embodiments may include programmed, programmable devices, or some combination thereof (e.g., PLAs, PLDs, ASICs, microcontroller, microprocessor), and may include one or more data stores (e.g., cell, register, block, page) that provide single or multi-level digital data storage capability, and which may be volatile, non-volatile, or some combination thereof. Some control functions may be implemented in hardware, software, firmware, or a combination of any of them.

[0123] Computer program products may contain a set of instructions that, when executed by a processor device, cause the processor to perform prescribed functions. These functions may be performed in conjunction with controlled devices in operable communication with the processor. Computer program products, which may include software, may be stored in a data store tangibly embedded on a storage medium, such as an electronic, magnetic, or rotating storage device, and may be fixed or removable (e.g., hard disk, floppy disk, thumb drive, CD, DVD).

[0124] Although an example of a system, which may be portable, has been described with reference to the above figures, other implementations may be deployed in other processing applications, such as desktop and networked environments.

[0125] Temporary auxiliary energy inputs may be received, for example, from chargeable or single use batteries, which may enable use in portable or remote applications. Some embodiments may operate with other DC voltage sources, such as a 9V (nominal) batteries, for example. Alternating current (AC) inputs, which may be provided, for example from a 50 / 60 Hz power port, or from a portable electric generator, may be received via a rectifier and appropriate scaling. Provision for AC (e.g., sine wave, square wave, triangular wave) inputs may include a line frequency transformer to provide voltage step-up, voltage step-down, and / or isolation.

[0126] Although particular features of an architecture have been described, other features may be incorporated to improve performance. For example, caching (e.g., LI, L2, . . .) techniques may be used. Random access memory may be included, for example, to provide scratch pad memory and or to load executable code or parameter information stored for use during runtime operations. Other hardware and software may be provided to perform operations, such as network or other communications using one or more protocols, wireless (e.g., infrared) communications, stored operational energy and power supplies (e.g., batteries), switching and / or linear power supply circuits, software maintenance (e.g., self-test, upgrades), and the like. One or more communication interfaces may be provided in support of data storage and related operations.

[0127] Some systems may be implemented as a computer system that can be used with various implementations. For example, various implementations may include digital circuitry, analog circuitry, computer hardware, firmware, software, or combinations thereof. Apparatus can be implemented in a computer program product tangibly embodied in an information carrier, e.g., in a machine-readable storage device, for execution by a programmable processor; and methods can be performed by a programmable processor executing a program of instructions to perform functions of various embodiments by operating on input data and generating an output. Various embodiments can be implemented advantageously in one or more computer programs that are executable on a programmable system including at least one programmable processor coupled to receive data and instructions from, and to transmit data and instructions to, a data storage system, at least one input device, and / or at least one output device. A computer program is a set of instructions that can be used, directly or indirectly, in a computer to perform a certain activity or bring about a certain result. A computer program can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0128] Suitable processors for the execution of a program of instructions include, by way of example, both general and special purpose microprocessors, which may include a single processor or one of multiple processors of any kind of computer. Generally, a processor will receive instructions and data from a read-only memory or a random-access memory or both. The essential elements of a computer are a processor for executing instructions and one or morememories for storing instructions and data. Generally, a computer will also include, or be operatively coupled to communicate with, one or more mass storage devices for storing data files; such devices include magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and optical disks. Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non-volatile memory, including, by way of example, semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, ASICs (application-specific integrated circuits).

[0129] In some implementations, each system may be programmed with the same or similar information and / or initialized with substantially identical information stored in volatile and / or non-volatile memory. For example, one data interface may be configured to perform auto configuration, auto download, and / or auto update functions when coupled to an appropriate host device, such as a desktop computer or a server.

[0130] In some implementations, one or more user-interface features may be custom configured to perform specific functions. Various embodiments may be implemented in a computer system that includes a graphical user interface and / or an Internet browser. To provide for interaction with a user, some implementations may be implemented on a computer having a display device. The display device may, for example, include an LED (light-emitting diode) display. In some implementations, a display device may, for example, include a CRT (cathode ray tube). In some implementations, a display device may include, for example, an LCD (liquid crystal display). A display device (e.g., monitor) may, for example, be used for displaying information to the user. Some implementations may, for example, include a keyboard and / or pointing device (e.g., mouse, trackpad, trackball joystick), such as by which the user can provide input to the computer.

[0131] In various implementations, the system may communicate using suitable communication methods, equipment, and techniques. For example, the system may communicate with compatible devices (e.g., devices capable of transferring data to and / or from the system) using point-to-point communication in which a message is transported directly from the source to the receiver over a dedicated physical link (e.g., fiber optic link, point-to-point wiring, daisy-chain). The components of the system may exchange information by any form or medium of analog ordigital data communication, including packet-based messages on a communication network. Examples of communication networks include, e.g., a LAN (local area network), a WAN (wide area network), MAN (metropolitan area network), wireless and / or optical networks, the computers and networks forming the Internet, or some combination thereof. Other implementations may transport messages by broadcasting to all or substantially all devices that are coupled together by a communication network, for example, by using omni-directional radio frequency (RF) signals. Still other implementations may transport messages characterized by high directivity, such as RF signals transmitted using directional (i.e., narrow beam) antennas or infrared signals that may optionally be used with focusing optics. Still other implementations are possible using appropriate interfaces and protocols such as, by way of example and not intended to be limiting, USB 2.0, Firewire, ATA / IDE, RS-232, RS-422, RS-485, 802.11 a / b / g, Wi-Fi, Ethernet, IrDA, FDDI (fiber distributed data interface), token-ring networks, multiplexing techniques based on frequency, time, or code division, or some combination thereof. Some implementations may optionally incorporate features such as error checking and correction (ECC) for data integrity, or security measures, such as encryption (e.g., WEP) and password protection.

[0132] In various embodiments, the computer system may include Internet of Things (loT) devices. loT devices may include objects embedded with electronics, software, sensors, actuators, and network connectivity which enable these objects to collect and exchange data. loT devices may be in-use with wired or wireless devices by sending data through an interface to another device. loT devices may collect useful data and then autonomously flow the data between other devices.

[0133] Various examples of modules may be implemented using circuitry, including various electronic hardware. By way of example and not limitation, the hardware may include transistors, resistors, capacitors, switches, integrated circuits, other modules, or some combination thereof. In various examples, the modules may include analog logic, digital logic, discrete components, traces and / or memory circuits fabricated on a silicon substrate including various integrated circuits (e.g., FPGAs, ASICs), or some combination thereof. In some embodiments, the module(s) may involve execution of preprogrammed instructions, software executed by a processor, or some combination thereof. For example, various modules may involve both hardware and software.

[0134] In an illustrative aspect, a radiation dosimetry system (e.g., as shown in Fig. 1) may include at least one processor, for example, processor 305 as shown in Fig. 4. The system may include at least one data store 330 communicably coupled to at least one processor 305. In some examples, data store 330 may include a dynamic dosimetry computation engine, RDCE 105, as depicted in Fig. 4. Dynamic dosimetry computation engine RDCE 105 may be configured such that, when at least one processor 305 executes the engine, automatic dynamic dosimetry operations are performed, for instance. The operations may include, in response to receiving an image feed of an environment including radiation source 120 and person 115 wearing visual indicium (e.g., one or more tags 125), determining from the image feed, by computer vision engine (e.g., CVE 130), a relative three-dimensional position of the user to radiation source 120, for example. The determination may be based on the size and orientation of the visual indicium in the image feed relative to the predetermined geometry, as an example. The system may generate a radiation dosage received by person 115 by applying a dosage prediction engine (e.g., DPE 140) to the relative position of the user over time, for instance. Some examples may utilize predetermined radiation dosage distributions stored in look-up table (LUT) (e.g., predetermined radiation exposure LUT 210) correlated with a three-dimensional position of radiation source 120, as illustrated in FIG. 2A.

[0135] As an example, the system may include the camera 110. The image feed may, by way of example and not limitation, be received from a single camera 110.

[0136] The dynamic dosimetry computation engine may be configured to determine radiation exposure of person 115 within the image feed based on an activity state of radiation source 120. The activity state may, for example, be determined from the image feed.

[0137] The operations may, for instance, further include generating a user body orientation model by applying a pose estimation model (e.g., of TPOIE 345) to the image feed and the predetermined geometry of the visual indicium The user body orientation model may, for example, include a relative location of multiple body portions of person 115. The operations may, for example, include generating the radiation dosage as a distributed radiation dosage by further applying the dosage prediction engine to the user body orientation model. The distributed radiation dosage may associate the body portions with corresponding measures of radiation exposure over time.

[0138] As an example, the system may include live visual display 170 operably coupled to at least one processor 305 such that the generated radiation dosage received by person 115 is displayed on live visual display 170.

[0139] Live visual display 170 may include a visual overlay on the image feed of a radiation distribution such that person 115 may visualize their position relative to the radiation distribution.

[0140] The predetermined radiation distribution may include a pre-computed radiation distribution, as shown in LUT 210. The visual overlay may, for example, be generated by retrieving the pre-computed radiation distribution for radiation source 120 and translating, rotating, and / or scaling this distribution based on the current geometry of radiation source 120.

[0141] A current geometry of radiation source 120 may be determined based on detecting the size and orientation in the image feed of a second visual indicium of predetermined geometry, as an example.

[0142] System 100 may include alarm 175 communicably coupled to at least one processor 305 and configured such that a real-time alarm is provided to person 115 in response to the generated radiation dosage received by person 115 exceeding at least one predetermined threshold.

[0143] As an example, the system may include radiation source 120. Radiation source 120 may, by way of example and not limitation, include a C-arm medical diagnostic radiography device, and the environment may include a medical operating room.

[0144] Predetermined radiation dosage distributions may be stored in LUT 210, as shown in FIG. 2A.

[0145] In an illustrative aspect, a hazard exposure dosimetry system may include, by way of example, at least one processor 305. The system may also include at least one data store 330 communicably coupled to processor 305. Data store 330 may include a dynamic hazard exposure dosimetry computation engine, RDHE 135. Computation engine RDHE 135 may be configured, for instance, to enable automatic dynamic hazard exposure dosimetry operations when executed by processor 305. Operations may include determining a relative three-dimensional position of the user to a hazard source from an image feed in response to receiving an image feed of an environment with a hazard source and person 115 wearing tags 125. This determination may be based, as an example, on the size and orientation of tags 125 in the image feed relative to their predetermined geometry, by computer vision engine CVE 130. The system may generate ahazard dosage received by person 115 by applying dosage prediction engine DPE 140 to the relative position of the user over time. This process may involve applying predetermined hazard dosage distributions stored in a LUT (e.g., predetermined radiation exposure LUT 210), correlated with a three-dimensional position of the hazard source.

[0146] The system may include camera 110 configured to generate the image feed, in some embodiments. The system may include a single camera, for instance.

[0147] As an example, the system may include the hazard source, which may include a C-arm medical diagnostic radiography device, and the environment may incorporate a medical operating room.

[0148] In an illustrative aspect, a computer-implemented method for hazard exposure dosimetry may incorporate various features. A processor 305 may, for example, receive an image feed from a camera 110. The image feed may include, for example, an environment containing a radiation source 120 and a person 115 who may wear visual indicium tags 125 of a predetermined geometry. The computer vision engine (e.g., CVE 130) may, for example, determine a relative three-dimensional position of the person 115 to the radiation source 120. This determination may be based on the size and orientation of the visual indicium tags 125 in the image feed, relative to their predetermined geometry. A dosage prediction engine (e.g., DPE 140) may be applied to the relative position of the person 115 over time. The DPE may utilize predetermined hazard dosage distributions correlated with the three-dimensional position of the radiation source 120. This process may, for example, generate a hazard dosage received by the person 115.

[0149] In an illustrative aspect, a computer program product may, for example, be encoded on a non-transitory machine-readable medium. This computer program product may, for example, include instructions that, when executed on processor 305, cause the processor to perform a method. The method may include receiving an image feed from camera 110 of an environment that includes a radiation source 120 and a person 115 wearing visual indicium tags 125 of predetermined geometry. The CVE 130 may determine, from the image feed, a relative three- dimensional position of the person 115 to the radiation source 120. This determination may be based on the size and orientation of the visual indicium tags 125 relative to their predetermined geometry. A DPE 140 may, for example, be used to generate a hazard dosage received by the person 115, by applying it to the relative position of the person 115 over time and predeterminedhazard dosage distributions correlated with the three-dimensional position of the radiation source 120.

[0150] As an example, the image feed may be received from a single camera (e.g., camera 110).

[0151] As an example, the dynamic hazard exposure dosimetry computation engine (e.g., RDHE 135) may be configured to determine hazard exposure of the person (e.g., Person 115) within the image feed based on an activity state of the hazard source (e.g., radiation source 120).

[0152] As an example, the activity state may be determined from the image feed (e.g., camera 110).

[0153] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined based on a change in brightness in the image feed.

[0154] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined based on identifying an illuminated indicator light on the radiation source.

[0155] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined based on detecting a shadow or outline of the radiation source changing in the image feed.

[0156] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined by detecting user interaction with control elements associated with the radiation source in the image feed.

[0157] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined based on visual indicium attached to the radiation source indicating its operational status.

[0158] As an example, the activity state of the radiation source (e.g., radiation source 120) may be determined based on detecting movement in the image feed associated with activation of the radiation source.

[0159] As an example, the operations may further include generating a user body orientation model by applying a pose estimation model to the image feed and the predetermined geometry of the visual indicium (e.g., Tags 125), the user body orientation model may include a relative location of a plurality of body portions of the user (e.g., Person 115); and generating the hazard dosage as a distributed hazard dosage by further applying the dosage prediction engine (e.g., DPE 140) to the user body orientation model, the distributed hazard dosage associating the plurality of body portions with corresponding measures of hazard exposure over time.

[0160] As an example, the operations may further include operating a live visual display (e.g., live visual display 170), operably coupled to the at least one processor, such that the generated hazard dosage received by the person (e.g., person 115) is displayed on the live visual display.

[0161] As an example, the operations may generate a hazard visualization such that the live visual display (e.g., live visual display 170) shows a visual overlay on the image feed of a hazard distribution such that the person (e.g., person 115) can visualize their position relative to the hazard distribution.

[0162] As an example, the visual overlay may be generated by retrieving a pre-computed hazard distribution for the hazard source (e.g., radiation source 120), and translating, rotating, and scaling the pre-computed hazard distribution based on a current geometry of the hazard source.

[0163] As an example, a current geometry of the hazard source (e.g., radiation source 120) may be determined based on detecting a size and orientation in the image feed (e.g., camera 110) of a second visual indicium of predetermined geometry.

[0164] As an example, the system may further include an alarm (e.g., live hazard alarm 175) communi cably coupled to the at least one processor and configured such that a real-time alarm is provided to the person (e.g., person 115) in response to the generated hazard dosage received by the person exceeding at least one predetermined threshold.

[0165] As an example, the predetermined hazard dosage distributions may be stored in a look-up table (e g., predetermined radiation exposure LUT 210).

[0166] As an example, the at least one data store may further include a predictive computation engine configured such that, when the at least one processor executes the predictive computation engine, anticipatory dosimetry operations are performed. The operations may include, in response to the relative three-dimensional position of the user (e.g., person 115) to the hazard source (e.g., radiation source 120), applying a prediction model (e.g., DPE 200) incorporating historical hazard exposure trends to the relative position of the user to anticipate future exposure scenarios; and generating recommendations for user actions to mitigate future radiation exposure based on the anticipated scenarios and storing these recommendations in the data store for user reference.

[0167] As an example, the hazard may include ionizing radiation, the hazard source may include a radiation source (e.g., radiation source 120), and hazard dosage may include a radiation dosage.

[0168] As an example, the hazard may include ultraviolet light, the hazard source may include an ultraviolet light source (e.g., radiation source 120), and hazard dosage may include an ultraviolet light dosage.

[0169] As an example, the hazard may include sound above a predetermined decibel level, the hazard source may include an acoustic source (e.g., radiation source 120), and hazard dosage may include an acoustic exposure dosage.

[0170] As an example, the hazard may include chemical exposure, the hazard source may include a chemical source (e.g., radiation source 120), and hazard dosage may include a chemical exposure dosage.

[0171] In an illustrative aspect, a real-time visual overlay system may include at least one processor; and at least one data store communicably coupled to the at least one processor and may include a dynamic overlay computation engine configured such that, when the at least one processor executes the dynamic overlay computation engine, automatic visual overlay operations are performed. The operations may include, in response to receiving an image feed of an environment including a source (e.g., radiation source 120) and a person (e.g., Person 115) wearing a visual indicium of predetermined geometry (e.g., Tags 125), determining from the image feed, by a computer vision engine (e.g., CVE 130), a relative three-dimensional position of the user to the source based on a size and orientation of the visual indicium in the image feed relative to the predetermined geometry; and generating a real-time visual overlay representing the relative position and radiation exposure levels of the user by applying a visualization engine to the relative position of the user and predetermined data distributions correlated with a three- dimensional position of the source.

[0172] In an illustrative aspect, a method for real-time exposure visualization may include receiving, by a processor, an image feed of an environment including a radiation source (e.g., radiation source 120) and a person (e.g., Person 115) wearing a visual indicium of predetermined geometry (e.g., Tags 125); determining, by a computer vision engine (e.g., CVE 130), a relative three-dimensional position of the person to the radiation source from the image feed based on a size and orientation of the visual indicium in the image feed relative to the predetermined geometry; generating, by a visualization engine, a real-time visual overlay (e.g., time-based exposure report 180) displaying radiation exposure levels based on the relative position of the person and stored radiation dosage data associated with the radiation source; and providing, by alive feedback system, the visual overlay to a display device for the person to visualize real-time exposure levels.

[0173] A number of implementations have been described. Nevertheless, it will be understood that various modifications may be made. For example, advantageous results may be achieved if the steps of the disclosed techniques were performed in a different sequence, or if components of the disclosed systems were combined in a different manner, or if the components were supplemented with other components. Accordingly, other implementations are contemplated within the scope of the following claims, the original claims being hereby specifically incorporated into the specification.

Claims

Claims1. A radiation dosimetry system comprising: at least one processor (305); at least one data store (325, 330) communicably coupled to the at least one processor (305) and comprising a dynamic dosimetry computation engine (105) configured such that, when the at least one processor (305) executes the dynamic dosimetry computation engine (105), automatic dynamic dosimetry operations are performed, the operations comprising: in response to receiving an image feed of an environment comprising a radiation source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry, determine from the image feed, by a computer vision engine (130), a relative three-dimensional position of the person (115) to the radiation source (120) based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; and generate a radiation dosage received by the person (115) by applying a dosage prediction engine (140) to the relative three-dimensional position of the person over time and predetermined radiation dosage distributions correlated with a three- dimensional position of the radiation source (120).

2. The system of claim 1, further including a single camera (110), wherein the image feed is received from the single camera (110).

3. The system of any one of claims 1-2, wherein the dynamic dosimetry computation engine (105) is configured to determine radiation exposure of the person (115) within the image feed based on an activity state of the radiation source (120).

4. The system of claim 3, wherein the activity state is determined from the image feed.

5. The system of any one of claims 1-4, the operations further including: generate a user body orientation model by applying a pose estimation model (345) to the image feed and the predetermined geometry of the visual indicium (125), the user body orientation model including a relative location of a plurality of body portions of the person; and generate the radiation dosage as a distributed radiation dosage by further applying the dosage prediction engine (140) to the user body orientation model, the distributed radiationdosage associating the plurality of body portions with corresponding measures of radiation exposure over time.

6. The system of any one of claims 1-5, further including: a live visual display (170) operably coupled to the at least one processor such that the generated radiation dosage received by the person (11 ) is displayed on the live visual display (170).

7. The system of claim 6, wherein the live visual display (170) includes a visual overlay (430) on the image feed of a radiation distribution, allowing the person (115) to visualize their position relative to the radiation distribution.

8. The system of claim 7, wherein the predetermined radiation distribution includes a precomputed radiation distribution, and the visual overlay is generated by retrieving the precomputed radiation distribution for the radiation source (120), and translating, rotating, and scaling the pre-computed radiation distribution based on a current geometry of the radiation source (120).

9. The system of any one of claims 1-8, wherein a current geometry of the radiation source (120) is determined based on detecting a size and orientation in the image feed of a second visual indicium (125) of predetermined geometry.

10. The system of any one of claims 1-9, further including an alarm communicably coupled to the at least one processor and configured such that a real-time alarm (175) is provided to the person (115) in response to the generated radiation dosage received by the person (115) exceeding at least one predetermined threshold.

11. The system of any one of claims 1-10, wherein: the system includes the radiation source (120), which radiation source (120) includes a C-arm medical diagnostic radiography device, and the environment includes a medical operating room.

12. The system of any one of claims 1-11, wherein the predetermined radiation dosage distributions are stored in a look-up table (210).

13. A hazard exposure dosimetry system comprising: at least one processor (305); at least one data store (325, 330) communicably coupled to the at least one processor and comprising a dynamic hazard exposure dosimetry computation engine (105) configured such that, when the at least one processor executes the dynamic hazard exposure dosimetry computation engine, automatic dynamic hazard exposure dosimetry operations are performed, the operations comprising: in response to receiving an image feed of an environment comprising a hazard source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry, determine from the image feed, by a computer vision engine (130), a relative three-dimensional position of the person (115) to the hazard source (120) based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; and generate a hazard dosage received by the person (115) by applying a dosage prediction engine (140) to the relative three-dimensional position of the person (115) over time and predetermined hazard dosage distributions correlated with a three- dimensional position of the hazard source (120).

14. The system of claim 13, further comprising a camera (110) configured to generate the image feed.

15. The system of any one of claims 13-14, wherein: the system comprises the hazard source (120), which hazard source comprises a C-arm medical diagnostic radiography device, and the environment comprises a medical operating room.

6. A computer-implemented method, comprising: receiving, by at least one processor (305), an image feed of an environment comprising a hazard source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry; determining, from the image feed, by a computer vision engine (130), a relative three- dimensional position of the person (115) to the hazard source (120) based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; and generating, by applying a dosage prediction engine (140) to the relative three- dimensional position of the person (115) over time and predetermined hazard dosage distributions correlated with a three-dimensional position of the hazard source (120), a hazard dosage received by the person (115).

17. A computer program product encoded on a non-transitory machine-readable medium (325, 330), the computer program product comprising instructions that, when executed on at least one processor (305), cause the processor to perform a method comprising: receiving an image feed of an environment comprising a hazard source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry; determining, by a computer vision engine (130), a relative three-dimensional position of the person (115) to the hazard source (120) from the image feed based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; and generating a hazard dosage received by the person (115) by applying a dynamic hazard exposure dosimetry computation engine (105) comprising a dosage prediction engine (140) to the relative three-dimensional position of the person (115) over time and predetermined hazard dosage distributions (210) correlated with a three-dimensional position of the hazard source (120).

18. Any one of claims 1-17, wherein the image feed is received from a single camera (110).

19. Any one of claims 13-18 wherein the dynamic hazard exposure dosimetry computation engine is configured to determine hazard exposure of the person (115) within the image feed based on an activity state of the hazard source (120).

20. Claim 19, wherein the activity state is determined from the image feed.

21. Claim 20, wherein the activity state of the radiation source (120) is determined based on a change in brightness in the image feed.

22. Any one of claims 20-21, wherein the activity state of the radiation source (120) is determined based on identifying an illuminated indicator light on the radiation source (120).

23. Any one of claims 20-22, wherein the activity state of the radiation source (120) is determined based on detecting a shadow or outline of the radiation source (120) changing in the image feed.

24. Any one of claims 20-23, wherein the activity state of the radiation source (120) is determined by detecting user interaction with control elements associated with the radiation source (120) in the image feed.

25. Any one of claims 20-24, wherein the activity state of the radiation source (120) is determined based on visual indicium (125) attached to the radiation source (120) indicating its operational status.

26. Any one of claims 20-25, wherein the activity state of the radiation source (120) is determined based on detecting movement in the image feed associated with activation of the radiation source (120).

27. Any one of claims 20-26, the operations further comprising: generate a user body orientation model by applying a pose estimation model (345) to the image feed and the predetermined geometry of the visual indicium (125), the user body orientation model comprising a relative location of a plurality of body portions of the person (115); and generate the hazard dosage as a distributed hazard dosage by further applying the dosage prediction engine (140) to the user body orientation model, the distributed hazard dosage associating the plurality of body portions with corresponding measures of hazard exposure over time.

28. Any one of claims 13-27, the operations further comprising operating a live visual display (170), operably coupled to the at least one processor, such that the generated hazard dosage received by the person (115) is displayed on the live visual display (170).

29. Claim 28, the operations generating a hazard visualization such that the live visual display (170) shows a visual overlay on the image feed of a hazard distribution such that the person (115) can visualize their position relative to the hazard distribution.

30. Claim 29, wherein the visual overlay is generated by retrieving a pre-computed hazard distribution for the hazard source (120), and translating, rotating, and scaling the pre-computed hazard distribution based on a current geometry of the hazard source (120).

31. Any one of claims 13-30, wherein a current geometry of the hazard source (120) is determined based on detecting a size and orientation in the image feed of a second visual indicium (420) of predetermined geometry.

32. Any one of claims 13-31, further comprising an alarm communicably coupled to the at least one processor and configured such that a real-time alarm is provided to the person (115) in response to the generated hazard dosage received by the person (115) exceeding at least one predetermined threshold.

33. Any one of claims 13-32, wherein the predetermined hazard dosage distributions are stored in a look-up table (210).

34. Any one of claims 13-33, wherein the at least one data store further comprises a predictive computation engine (200) configured such that, when the at least one processor executes the predictive computation engine, anticipatory dosimetry operations are performed, the operations comprising: in response to the relative three-dimensional position of the person (115) to the hazard source (120); apply a prediction model (145) incorporating historical hazard exposure trends to the relative three-dimensional position of the person (115) to anticipate future exposure scenarios; and generate recommendations for user actions to mitigate future radiation exposure based on the anticipated scenarios and store these recommendations in the data store for user reference.

35. Any one of claims 13-34, wherein: the hazard comprises ionizing radiation, the hazard source comprises a radiation source (120), and hazard dosage comprises a radiation dosage.

36. Any one of claims 13-35, wherein: the hazard comprises ultraviolet light, the hazard source comprises an ultraviolet light source, and hazard dosage comprises an ultraviolet light dosage.

37. Any one of claims 13-36, wherein: the hazard comprises sound above a predetermined decibel level, the hazard source comprises an acoustic source, and hazard dosage comprises an acoustic exposure dosage.

38. Any one of claims 13-37, wherein: the hazard comprises chemical exposure, the hazard source comprises a chemical source, and hazard dosage comprises a chemical exposure dosage.

9. Any one of claims 13-38, the operations further comprising: determine the three-dimensional position of the person (115) relative to the hazard source (120) corresponds to a level of hazard dosage, in the predetermined hazard dosage distribution, higher than a hazard level threshold; in response to determining the position of the person (115) is higher than the hazard level threshold, generate and transmit a signal such that a wearable device (1110) associated with the person (115) generates a notification corresponding to the hazard level.

40. A real-time visual overlay system comprising: at least one processor (305); at least one data store (325, 330) communicably coupled to the at least one processor and comprising a dynamic overlay computation engine (105) configured such that, when the at least one processor executes the dynamic overlay computation engine, automatic visual overlay operations are performed, the operations comprising: in response to receiving an image feed of an environment comprising a source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry, determine from the image feed, by a computer vision engine (130), a relative three- dimensional position of the person (115) to the source (120) based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; and generate a real-time visual overlay (430) representing the relative three- dimensional position and radiation exposure levels of the person (115) by applying a dosimetry hazard engine (135) to the relative three-dimensional position of the person (115) and predetermined data distributions correlated with a three-dimensional position of the source (120).

41. The system of claim 40 according to any of claims 1-39.

42. A method for real-time exposure visualization comprising: receiving, by a processor, an image feed of an environment comprising a radiation source (120) and a person (115) wearing a visual indicium (125) of predetermined geometry; determining, by a computer vision engine (130), a relative three-dimensional position of the person (115) to the radiation source (120) from the image feed based on a size and orientation of the visual indicium (125) in the image feed relative to the predetermined geometry; generating, by a dosimetry hazard engine (135), a real-time visual overlay displaying radiation exposure levels based on the relative three-dimensional position of the person (115) and stored radiation dosage data associated with the radiation source (120); and providing, by a live feedback system (170, 175), the visual overlay to a display device for the person (115) to visualize real-time exposure levels.

43. The system of claim 42 according to any of claims 1-41.

Images