Instrument parameter determination based on Sample Tube Identification
A system and method for reducing the responsibility of the user significantly by applying an optical system that can identify container like sample tubes with respect to their characteristics, e.g., shapes and inner dimensions, from their visual properties by capturing images from a rack comprising container and processing said images for reliably identifying a container tyle.
1 . A method for determining hematocrit, the method comprising
receiving at least one image of a container comprising a centrifuged blood sample using a camera,
processing the at least one image to determine boundaries of separate layers of material in the centrifuged blood sample, and
determining hematocrit based on one or more of layers, liquid levels, or liquid volumes percentage of a first layer and a second layer in the container based on the boundaries determined from the at least one image.
2 . The method of claim 1 , further comprising capturing a plurality of images of the container using a camera and creating a segmented picture comprising stacked segments from the plurality of images.
3 . The method of claim 2 , wherein the at least one image comprises the segmented picture.
4 . The method of claim 2 , wherein each of the plurality of images are captured at different positions.
5 . The method of claim 1 , wherein capturing at least one image of a container comprises capturing an image of a rack including the at least the container.
6 . The method of claim 5 , further comprising capturing at least a second image of the rack, measuring height, width, or both of the container at different heights in each image, and calculating mean values with standard deviation for determining the container's dimensions.
7 . The method of claim 5 , further comprising illuminating the container using a light source positioned to illuminate a first side of the rack opposite a second side of the rack, wherein the sensor is arranged to capture an image of the second side of the rack.
8 . The method of claim 1 , further comprising processing the at least one image to determine one or more characteristics of the container selected from the group consisting of width, height, shape, and presence of a cap.
9 . The method of claim 1 , whether comprising processing the at least one image to determine presence or absence of a false bottom of the container.
10 . The method of claim 1 , further comprising determining a type or class of the container based at least on the one or more characteristics of the container.
11 . The method according to claim 10 , wherein determining the type or class of the container based on the one or more characteristics of the container comprises identifying the type or class of the container using container data stored in a database, and wherein the database further comprises a set of instrument parameters assigned to a container type or container class.
12 . The method of claim 1 , further comprising wherein the first layer comprises red blood cells and the second layer comprises plasma.
13 . The method of claim 11 , comprising the step of capturing multiple images of the container during rotation of the container in front of the sensor and forming a segmented picture from the multiple images.
14 . The method of claim 11 , further comprising the step of applying the segmented picture to a convolutional neural network (CNN) for determining an upper boundary and a lower boundary of a plasma layer in the segmented picture for generating a bounding box enclosing the plasma layer in all segments of the segmented picture.
15 . The method of claim 14 , comprising the step of rearranging the segments of the segmented picture prior to determining again the upper boundary and the lower boundary of the plasma layer in the newly arranged segmented picture for generating a bounding box enclosing the plasma layer in all segments of the segmented picture.
16 . The method of claim 1 , further comprising:
determining a region of interest (ROI) in the at least one image by identifying reference points;
determining an upper end of the rack in the ROI; and
determining edges of an upper end of the container within the ROI.
17 . The method of claim 1 , wherein determining presence or absence of the false bottom of the container further comprises capturing a plurality of images of the container while rotating the container, and processing the plurality of images to determine presence or absence of the false bottom.
18 . A system comprising,
a designated position to receive a container,
a camera to capture one or more images of the container in the designated position, and
a convolutional neural network (CNN) to receive the one or more images of the container, the CNN for determining boundaries of separated layers of a material located in the container for determining hematocrit based on one or more of layers, liquid levels or liquid volumes percentage based on the boundaries determined from the one or more images.
19 . The system of claim 18 , wherein the CNN additionally determines the upper and the lower boundary of the plasma layer in the segmented picture for generating a bounding box enclosing the plasma layer in all segments of the segmented image.
20 . The system of claim 18 , further comprising a rotator to rotate the container in front of the camera.